In the last few years the traffic on roads has increased by leaps and bounds and so have the problems of over-speeding, car theft and jumping the red light. Due to the above-mentioned problems, vehicle tracking, recognition and speed detection have gained immense importance in modern traffic control systems.

 Previous techniques required special hardware, calibration of hardware or failed to identify the vehicle. The proposed goal of the thesis is to develop real time system that will recognize license plate of the vehicle and its movement parameters like position, speed and acceleration from video sequences. Each vehicle has a unique license plate number. License plate recognition system makes use ofthis unique number for variety of applications such as border monitoring, parking management, toll management, car ownership etc. Moreover, speed estimation of vehicles from videos has gained importance recently as it can be used in detecting over-speeding cars, automatic garages and toll gates operation, automated driving etc.

Tasks to be performed by the student include:

·         Develop an algorithm for licence plate region detection and license plate number identification on real-time videos;

·         Teach an artificial neural network to find the license plates on images and extract their unique license plate identifiers;  

·         Estimate car position, speed, acceleration based on the license plate’s area change and right/left shift;

·         Integrate the developed algorithm into an existing Internet of Things application;

·         Calculate additional information based on the computed metrics and use this information to develop new functionalities for the application;

·         Perform tests and evaluation of performance on the developed system;